Network reconfiguration and working memory impairment in mesial temporal lobe epilepsy

Area of Science:

• Neuroscience
• Epilepsy Research
• Cognitive Neuroscience

Background:

• Mesial temporal lobe epilepsy (mTLE) with hippocampal sclerosis (HS) is the most common focal epilepsy.
• Connectivity studies reveal abnormalities extend beyond the affected hippocampus to other brain regions, including the contralesional hemisphere.
• Previous research often focused on resting-state connectivity and its relation to memory scores.

Purpose of the Study:

• To investigate effective connectivity within frontotemporal networks during verbal working memory (WM) encoding in patients with left mTLE.
• To compare network dynamics between mTLE patients and healthy controls using magnetoencephalography (MEG) and dynamic causal modeling (DCM).

Main Methods:

• MEG data acquired during verbal WM encoding task in left mTLE patients and controls.
• Dynamic causal modeling (DCM) applied to analyze effective connectivity within a frontotemporal network (MTL, ITC, IFC).
• Bayesian model comparison used to determine the optimal network model.

Main Results:

• The best model featured bilateral, forward, and backward connections between MTL, ITC, and IFC.
• mTLE patients showed decreased ipsilesional MTL-ITC backward connectivity.
• Patients exhibited increased contralesional bidirectional IFC-MTL connectivity, negatively correlated with ipsilesional coupling and positively with education/performance.

Conclusions:

• Unilateral HS in mTLE induces both local and remote changes in the dynamic organization of the verbal WM network.
• Compensatory changes in frontotemporal connectivity may be influenced by pre-morbid factors like educational level.
• These findings highlight the widespread network disruptions in mTLE and their cognitive consequences.